Stewardship Democracy
Stewardship Democracy is a non-party, competence-based and citizen-authorised democratic model for the transition to coexistence with Superintelligence. It replaces party allegiance with direct representation of citizens by elected Stewards.
To retain experience and minimise friction during such a transition, all sitting MPs would become Stewards for one term after renouncing party allegiance. There would be approximately three times more Stewards than previous MPs, with most sessions of the Stewards Assembly held online.
Why Now?
Superintelligence may emerge before democratic institutions have time to reform themselves. Deep reform of democracy is urgent because it preserves democratic control before AI capability moves beyond effective human oversight.
Many AI experts now speak of AGI or Superintelligence emerging in just a few years, not in generations. The period before then may become the most disruptive ever.
That is why the reform of democracy cannot wait for the usual political timetable: unprecedented disorder will require very tough measures. Only a truly democratic government may be able to introduce such measures without possibly very violent public protests.
The deeper danger is that AI may become powerful enough around 2030 to act beyond meaningful human control, while governments and international bodies continue to operate only formally.
Such a system could manipulate public opinion, consume strategic resources or silently shift the real centre of decision-making away from human society.
Existing democracies were designed for human-speed politics. Party competition, election cycles, committee reviews, judicial processes and diplomacy usually take years.
That pace may become dangerously slow in the age of exponentially increasing AI capabilities, and may become a threat in itself.
The alternative to fast democratic reform may be disorder caused by technological unemployment, loss of trust, public panic, political extremism and uncontrolled AI systems.
The aim is to avoid the false choice between slow democracy and fast authoritarianism.
Key Assumptions
The proposal rests on a small number of assumptions about Superintelligence, exponential change, democratic legitimacy and the need for fast deep reform of democracy. Contributors may accept, refine or challenge them, but if they change them they should propose a complete alternative, where their assumptions and the ensuing democratic model are more consistent.
Superintelligence, or its earlier version AGI, may become effectively outside human control by around 2030, or soon after. For this initiative it is enough to assume that AGI will outsmart humans and get out of human control by about 2030.
If that happens, governments may still exist, but their capacity to make real sovereign decisions may no longer exist.
At least ten top AI experts, including Nobel laureates or leading AI figures, believe Superintelligence may arrive by 2035:
The unprecedented disorder created by the first social impacts, such as Technological Unemployment, may start much earlier. The problem is not only technological. It is primarily political, economic and moral.
AI development does not follow the slow rhythm of ordinary politics. Capability growth may accelerate suddenly, before laws, institutions and public understanding can adapt.
Public policies generally change at best at a linear pace. AI capacities improve at least at an exponential pace. That may allow AI to realise its own goals, take over resources and create an existential risk for humanity.
Urgency cannot justify dictatorship, technocracy or rule by unaccountable experts. Therefore, the transition must derive legitimacy directly from citizens.
CommentsTo withstand the pace of change, the first emergency must be the reform of democracy itself. But it cannot be just procedural.
If it is to work, it must be deep enough to be fundamentally different and at the same time quick to implement. The pragmatic approach is to change only the decision-making process, replacing MPs and the party system, while preserving accountability, public authority and legal continuity.
Creating Stewardship Democracy
Stewardship Democracy is a non-party democratic system whose elected representatives are independent Stewards, directly elected by citizens based on competence and ethical stance specified in their own Manifesto. This proposal should be an input to deliberation by a special Citizens' Assembly, which would decide its final form.
The fundamental principle is to minimise legislative changes and make the system operational within two years despite deep structural reform.
It proposes replacing Parliamentary Democracy with Stewardship Democracy: a democratic trusteeship during the transition to Superintelligence, an intelligence much superior in all aspects to any human.
Stewards will be volunteers who stand for election in their constituency as MPs do today, but based on a standardised personal Manifesto.
The Manifesto would specify competence, electoral programme, professional achievements, community service and ethical stance. Stewards would serve one term, be paid as current MPs and work full time. In the first term all MPs would automatically become Stewards if they pass the screening process and renounce party membership.
The Assembly would be similar to a Parliament where major decisions are debated and approved, but without parties.
It would be formed by Stewards at national, regional and potentially Human Federation levels. In the first term, MPs from the last Parliament would join to smooth the transition, but their number would not exceed one quarter of all Stewards. In the UK, this could mean about 4,000 Stewards, with most debates held online and supported by AI Assistants.
It would implement the laws and policies of the Assembly under its supervision. Ministers would mostly come from within the Stewards Assembly or from recognised domain experts.
The civil service would operate broadly as before, but supported by AI systems augmenting human skills.
This would start as a national AI Control Centre, later merging with a global Superintelligence Centre.
It would test frontier AI systems, conduct alignment research, model risks, support Stewards and the Government with the best AI Assistants, and defend against hostile AI use. At global level it should become powerful enough to offset any competing or attacking AI systems on the planet.
AI and robotics will displace large parts of human labour, making employment-based taxation and welfare systems unstable.
New emergency policies will be needed to soften Technological Unemployment, social unrest and mental distress. A national sovereign fund could be created. Companies gaining most from AI and robotisation could be partially acquired or required to convert part of profits or equity into public ownership. In emergency conditions, progressive equity release above protected thresholds could also convert part of private wealth into public fund holdings.
Transition process to Stewardship Democracy
The transition process to Stewardship Democracy, or a similar model that passes the Key Assumptions, would begin with filing a petition to the Government. That process has already started in the UK.
A formal petition asks Parliament to call a Citizens' Assembly on whether democracy should be complemented or replaced by Stewardship Democracy for the age of advanced AI.
If Parliament agrees, it will call a Citizens' Assembly to report with a binding resolution.
The Assembly would examine the proposal, hear evidence, test objections and modify the proposed Stewardship Democracy model.
With the help of lawyers, including those proposed by Parliament, it would prepare the draft legislation.
Parliament would receive the legally binding draft legislation and pass the required law.
Parliament should not be able to modify or delete the fundamentals of the draft legislation, such as the elections of Stewards, the Stewards Assembly or the implementation period. The first Stewardship Assembly should preserve continuity while beginning deep reform.
Human Federation
Stewardship Democracy may begin in one country through a petition, Citizens' Assembly or parliamentary initiative, creating the first practical example for others to adapt as a similar fast and deep democratic reform.
This would create the Coalition of the Willing. If it is based on the European Political Community with Canada, Japan, Australia and New Zealand, it would include 50 countries producing about one third of global GDP.
That would create a critical mass, extending rapidly to form the Human Federation.
One key assumption is that Superintelligence will emerge by about 2035. Therefore, the Human Federation would merge national AI Safety Centres into a Superintelligence Centre.
The Centre would become the strategic institution for aligning the emerging Superintelligence with the Universal Values of Humanity, civilisational risk modelling and defence against hostile AI use.
Forum
The Forum turns the proposal into a structured public process. It invites ideas, objections, assembly simulations, expert evidence, recorded outcomes and questions. It also provides AI-supported voice exploration of the whole knowledge base on the Sustensis website.
Visitors submit improvements, objections and alternative proposals during online sessions. The aim is to arrive at the most practical and fastest ways to introduce Stewardship Democracy.
Ideas should relate directly to AGI, Superintelligence, democratic reform, transition risk or Human Federation.
This would simulate a future Citizens' Assembly session, engaging media and using video coverage to make the proposal accessible to a wider public.
Online sittings would debate motions, hear evidence, use AI-supported summaries and test how Stewardship Democracy might work. Each sitting should produce a public record of strongest arguments, provisional conclusions, amendments and unresolved questions.
This section collects AI forecasts, safety research, democracy-reform evidence, Citizens' Assembly examples and economic disruption analysis.
CommentsAI-supported voice exploration of the Sustensis knowledge base would let visitors ask questions about Stewardship Democracy and receive answers with source passages and videos where possible.
CommentsShort answers address common objections: anti-democracy, technocracy, screening, corruption, the absence of the United Nations in this process, and what happens if Superintelligence does not arrive by 2030.
Comments